Update FG document (#540)

Author: yangxudong

docker/Dockerfile_paitf115

@@ -0,0 +1,36 @@
+FROM dsw-registry.cn-shanghai.cr.aliyuncs.com/pai/tensorflow-training:1.15PAI-cpu-py36-ubuntu18.04
+
+COPY docker/sources_18.04.list /etc/apt/sources.list
+
+# necessary for later commands to take effect
+RUN md5sum /etc/apt/sources.list
+
+RUN apt-get update
+RUN apt-get install apt-utils inetutils-ping wget curl telnet vim strace libpq-dev curl libsasl2-dev gcc g++ unzip openjdk-8-jdk -y
+
+RUN mkdir /EasyRec
+COPY requirements /EasyRec/requirements
+COPY requirements.txt /EasyRec/
+COPY easy_rec /EasyRec/easy_rec/
+COPY setup.cfg  /EasyRec/
+COPY setup.py /EasyRec/
+COPY MANIFEST.in /EasyRec/
+COPY README.md /EasyRec/
+COPY scripts /EasyRec/scripts
+
+RUN curl "http://easyrec.oss-cn-beijing.aliyuncs.com/tools/odpscmd_public_0.45.0.zip" -o /EasyRec/odpscmd_public.zip
+RUN mkdir /usr/local/odps_clt/ && cd /usr/local/odps_clt/ && unzip /EasyRec/odpscmd_public.zip
+RUN ln -s /usr/local/odps_clt/bin/odpscmd /usr/local/bin/odpscmd
+RUN pip3 install --upgrade pip
+RUN pip3 install pystack-debugger idna kafka-python -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install -r /EasyRec/requirements/runtime.txt
+RUN pip3 install -r /EasyRec/requirements/extra.txt   -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install http://easyrec.oss-cn-beijing.aliyuncs.com/3rdparty/graphlearn-1.1.0-cp36-cp36m-linux_x86_64.whl -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install http://easyrec.oss-cn-beijing.aliyuncs.com/releases/pai_automl-0.0.1rc1-py3-none-any.whl -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install https://easyrec.oss-cn-beijing.aliyuncs.com/3rdparty/common_io-0.4.2%2Btunnel-py2.py3-none-any.whl -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install tensorflow_probability==0.8 -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN cd /EasyRec && pip install .
+RUN rm -rf /EasyRec
+RUN python -c "import easy_rec; easy_rec.help(); import pyhive; import datahub; import kafka"
+
+COPY docker/hadoop_env.sh /opt/hadoop_env.sh

docker/Dockerfile_tf211

@@ -0,0 +1,36 @@
+FROM dsw-registry.cn-shanghai.cr.aliyuncs.com/pai/tensorflow:2.11-cpu-py39-ubuntu20.04-1
+COPY docker/sources_20.04.list /etc/apt/sources.list
+
+# necessary for later commands to take effect
+RUN md5sum /etc/apt/sources.list
+
+RUN apt-get update
+RUN apt-get install apt-utils inetutils-ping wget curl telnet vim strace libpq-dev curl libsasl2-dev gcc g++ unzip openjdk-8-jdk -y
+
+RUN mkdir /EasyRec
+COPY requirements /EasyRec/requirements
+COPY requirements.txt /EasyRec/
+COPY easy_rec /EasyRec/easy_rec/
+COPY setup.cfg  /EasyRec/
+COPY setup.py /EasyRec/
+COPY MANIFEST.in /EasyRec/
+COPY README.md /EasyRec/
+COPY scripts /EasyRec/scripts
+
+RUN curl "http://easyrec.oss-cn-beijing.aliyuncs.com/tools/odpscmd_public_0.45.0.zip" -o /EasyRec/odpscmd_public.zip
+RUN mkdir /usr/local/odps_clt/ && cd /usr/local/odps_clt/ && unzip /EasyRec/odpscmd_public.zip
+RUN ln -s /usr/local/odps_clt/bin/odpscmd /usr/local/bin/odpscmd
+RUN python -m pip install --upgrade pip
+RUN pip3 install pystack-debugger idna kafka-python -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install -r /EasyRec/requirements/runtime.txt -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install -r /EasyRec/requirements/extra.txt -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install https://tzrec.oss-cn-beijing.aliyuncs.com/third_party/graphlearn-1.2.0-cp39-cp39-linux_x86_64.whl -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+# RUN pip3 install http://easyrec.oss-cn-beijing.aliyuncs.com/releases/pai_automl-0.0.1rc1-py3-none-any.whl -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN pip3 install tensorflow_probability==0.19.0 -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+#RUN pip3 install encodings
+RUN pip3 install https://easyrec.oss-cn-beijing.aliyuncs.com/3rdparty/common_io-0.4.j2%2Btunnel-py2.py3-none-any.whl -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com
+RUN cd /EasyRec && pip install .
+RUN rm -rf /EasyRec
+# RUN python -c "import easy_rec; easy_rec.help(); import pyhive; import datahub; import kafka"
+
+COPY docker/hadoop_env.sh /opt/hadoop_env.sh

docs/images/other/fg.svg

@@ -462,27 +462,6 @@ ExprFeature：表达式特征
     - 当前版本未定义"&","|"的符号优先级，建议使用括号保证优先级。
     - customized normalization: "tf.math.log1p(user_age) / 10.0"
 
-EmbeddingVariable
-----------------------------------------------------------------
-Key Value Hash, 减少hash冲突, 支持特征准入和特征淘汰。
-
-.. code:: protobuf
-
-  model_config {
-    model_class: "MultiTower"
-    ...
-    ev_params {
-      filter_freq: 2
-    }
-  }
-- 配置方式:
-   - feature_config单独配置ev_params
-   - model_config里面统一配置ev_params
-
-- ev_params : EVParams
-   - filter_freq: 频次过滤, 低频特征噪声大,过滤噪声让模型更鲁棒
-   - steps_to_live: 特征淘汰, 淘汰过期特征,防止模型过大
-- Note: 仅在安装PAI-TF/DeepRec时可用
 
 特征选择
 ----------------------------------------------------------------